Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
High-order Interactions Modeling for Interpretable Multi-Agent Q-Learning
Authors: Qinyu Xu, Yuanyang Zhu, Xuefei Wu, Chunlin Chen
NeurIPS 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments demonstrate that QCo Fr not only consistently achieves better performance but also provides interpretability that aligns with our theoretical analysis. In this section, we conduct experiments to evaluate QCo Fr on three challenging benchmarks over the Level Based foraging (LBF) [31], the Star Craft Multi-Agent Challenge (SMAC) [32] and SMACv2 [33]. |
| Researcher Affiliation | Academia | Qinyu Xu School of Management and Engineering Nanjing University EMAIL Yuanyang Zhu School of Information Management Nanjing University EMAIL Xuefei Wu School of Management and Engineering Nanjing University EMAIL Chunlin Chen School of Robotics and Automation Nanjing University EMAIL |
| Pseudocode | Yes | We summarize the full pseudo-code of QCo Fr in Appendix C.1. |
| Open Source Code | No | The source code of implementations is from https://github.com/oxwhirl/pymarl. We will make the code available in the near future. |
| Open Datasets | Yes | We conduct experiments to evaluate QCo Fr on three challenging benchmarks over the Level Based foraging (LBF) [31], the Star Craft Multi-Agent Challenge (SMAC) [32] and SMACv2 [33]. |
| Dataset Splits | Yes | The detailed hyperparameter settings of LBF are shown in Table 1. The detailed hyperparameter settings of SMAC are shown in Table 2. Both tables include 'Test Episodes 32'. |
| Hardware Specification | Yes | All scenarios are trained on a system equipped with an NVIDIA RTX 3080TI GPU and an Intel i9-12900k CPU, with training time ranging from 1 to 16 hours per scenario, depending on the task complexity and episode length. |
| Software Dependencies | No | All implementations of algorithms are conducted on Star Craft II unit micro-management tasks (SC2.4.10). We implement all experiments within the Py MARL framework 2. |
| Experiment Setup | Yes | The detailed hyperparameter settings of LBF are shown in Table 1. The detailed hyperparameter settings of SMAC are shown in Table 2. |